Running a weekly project review
Weekly reviews drift into status theatre. Each lead describes their project, everyone nods, and the meeting ends having found nothing that was not already known.
A delivery lead has a portfolio of active projects and half an hour. Most are fine. One or two are not, and the review exists to find those.
Reading every project equally is how the meeting runs long and still misses things.
Each project carries three scores: health, tasks and setup. Sorting on them turns a status round into an exceptions list.
A low setup score is worth treating as a separate signal. It usually means the project record was created in a hurry and never finished, which makes the health and task scores less trustworthy for that project. Fix the record first. Otherwise the review spends its time debating a number that was never properly configured.
Why one record instead of two
A project review reads delivery scores next to people facts. Who is allocated, who is away, what the hours say. The review gets shorter when those do not have to be gathered first, and the meeting spends its time on judgement instead of collection.
What is switched on
- Projects
- Health, tasks and setup scores, plus status, priority, milestones and resource counts.
- Project boards
- Per-project Kanban boards give the task counts behind the score.
- Meetings
- Last meeting age is on the project row, so a quiet project is visible without opening it.
- Daily recap and timesheets
- Hours logged this week against a project say whether anyone actually worked on it.
- Resource allocation
- A project with health problems and no allocated hours has a simple diagnosis.
- Clients
- The client's local time matters when the review produces a call.
What the week looks like
Before the meeting
Sort the portfolio by health. Needs-attention projects come to the top.
Before the meeting
Scan setup scores. A low one usually means the project was never finished being configured.
In the meeting
Spend the time on exceptions. Healthy projects get named and skipped.
In the meeting
Open the board for anything contested. Task counts settle disagreements faster than opinions.
After
Note what changes and who owns it. Todo lists are shared, so an owner is a record.
Next week
Start from the same sort. The pattern over weeks is more useful than any single reading.
Questions people ask
Where do the health scores come from?
Project health is score-configurable in the organisation settings, with several keys under a project health category. That means the definition is yours rather than a vendor's opinion.
How do I compare tasks across projects?
On the portfolio view. Every board's progress rolls into its project's Tasks score, so all projects sit side by side.
What does a low setup score usually mean?
Usually that the project record was created quickly and never completed. It is a data-quality signal more than a delivery one, and it is worth fixing before trusting the other scores.